Two neighbors watch the same public decision. One sees protection for a vulnerable person; the other sees a different person put at risk. They may share a concern about suffering while disagreeing about who is likely to suffer, what causes it and whose account to trust. A disagreement can therefore be both deeply moral and based on different descriptions of the same event.
Kurt Gray and Samuel Pratt’s 2025 review, Morality in Our Mind and Across Cultures and Politics, argues that perceived harm is a common thread in moral judgment. It proposes that shared cognitive ingredients, including intention, causation and suffering, are combined with cultural interpretations to produce diverse moral values. This guide explains that argument, the competing pluralist account and the limits of using personal stories to bridge political divides. It describes a psychological theory of judgments, not a rule for deciding which side of a dispute is right.
The apparent puzzle: difference and similarity
People and cultures condemn different acts. One community may see gossip as a serious threat to social life; another may treat it as routine. Political groups may disagree sharply about a law because they imagine different victims and future consequences. Yet experiments also find recurring attention to whether an act was intentional, caused an outcome and involved suffering. Gray and Pratt ask how this common cognitive vocabulary coexists with different moral conclusions.
Their answer centers on perception. The same act can be understood as harmless by one person and as a serious threat by another. In this framework, “harm” is not limited to an immediately visible injury. People may perceive harm to bodies, relationships, dignity, institutions, future generations or beings they regard as vulnerable. Culture and experience influence which possible harms are noticed and how likely they seem.
This proposal is powerful partly because it can be misunderstood. Saying people perceive harm does not prove that every feared harm is real. Nor does it prove that harm is the only morally relevant consideration in philosophy or law. The review is a theory about how human minds construct moral judgments. Evidence about the actual consequences of a policy remains necessary.
Intention, causation and suffering
Consider an accidental injury and a deliberate attempt that fails. The outcome is worse in the first case, while the intention is worse in the second. Many people weigh both. The question “How wrong was the person?” may produce a different response from “How much repair is needed?” Gray and Pratt treat intention, causal responsibility and the experience of a possible victim as recurring elements in moral cognition.
Causation can itself be contested. One neighbor may say a policy directly prevents harm; another may emphasize indirect costs that appear years later. Both may use the language of protection while holding different models of how the world works. A conversation that asks only “Do you care about harm?” will miss the disputed causal chain.
Suffering also depends on whose experience is represented. A person may believe an act harms a child, a worker, a patient or a community that another person barely notices. Gray and Pratt use this variation to explain why distinct cultural values might arise from shared mental ingredients. This does not mean every concern can be reduced to one literal injured individual. It means perceived vulnerability can organize judgments in many domains.
Experiments on apparently “harmless” wrongs illustrate the debate. Gray, Schein and Ward found that some participants inferred victims or suffering even in scenarios designed to be harmless. That result challenges the assumption that participants’ condemnation occurred without any perceived harm. But it does not show that all possible cases of purity or convention are secretly harm judgments. Stimulus wording, social context and the breadth of the harm question matter.
Purity, loyalty and other moral vocabularies
Moral foundations theory offers a competing account. It proposes partly distinct moral concerns, including care, fairness, loyalty, authority and purity. Graham, Haidt and Nosek found political differences in the relative relevance of these concerns across several measures in their 2009 studies. On this view, a dispute about purity or loyalty need not be a disguised dispute about harm.
Gray and Pratt argue instead that people may connect purity, loyalty and authority to perceived harms. A breach of a ritual may be understood as threatening a community; disloyalty may be seen as exposing a group to danger; disrespect for an authority may be seen as undermining a protective institution. The theory is called constructionist because meanings emerge from basic ingredients assembled in context, rather than from a separate fixed mental module for every value.
Both accounts deserve careful testing. A pluralist questionnaire can show that people endorse different moral terms, but it may not show how each judgment was generated. A harm-centered explanation can illuminate hidden victims, but it risks becoming too flexible if every possible objection is redescribed as a harm after the fact. A strong test would specify in advance what each account predicts, use diverse examples and ask participants what they believe the consequences are.
Curry and colleagues’ cross-cultural study of morality as cooperation in 60 societies offers another reminder that the field has more than two lenses. Cooperation, kinship, reciprocity and property can also organize moral descriptions. No one framework should be treated as a complete map of every person’s ethical life merely because it explains a wide range of judgments.
Modularity and constructionism
In a modular picture, relatively distinct psychological systems are linked to different kinds of moral concern. The advantage is specificity: it can make predictions about which cues trigger which reactions. The difficulty is showing that the apparent categories arise from separate mechanisms rather than from shared processes responding to different contexts.
A constructionist picture begins with more general ingredients and asks how a situation, culture and learning combine them. The same cognitive capacities for tracking agents, intentions and distress can produce different judgments when the perceived facts change. Its advantage is that it connects cultural variation to common psychological processes. Its challenge is to make sufficiently precise predictions to be falsifiable.
Gray and Pratt favor the constructionist, harm-based account. This article does not treat that preference as a settled consensus. One can accept that perceived harm matters widely while still asking whether fairness, duty, sacredness or rights sometimes operate independently in a given judgment. The empirical issue is not decided by defining those concerns as harm in advance.
The distinction between descriptive and normative inquiry remains essential. Psychology can study why someone condemns an act. It cannot establish that the condemnation is justified simply because it feels like harm. A policy may really injure a group even if many observers fail to perceive it; a vivid imagined injury may be unsupported by evidence. Moral conversation needs both empathy and factual assessment.
How culture shapes what harm looks like
A cultural practice gives people concepts, examples and expectations. It can teach which relationships carry obligations, which agents can be vulnerable and which actions have long-range effects. A person who has lived through exclusion may perceive a threat in a phrase that another hears as harmless. A person who depends on a local institution may see a reform as jeopardizing a source of safety that outsiders overlook.
These interpretations are not random. They arise from histories, material conditions, stories and social learning. A cross-cultural comparison that translates only the word “harm” may miss how participants understand obligation, kinship or spiritual risk. Researchers need to learn the local account before deciding that two answers differ in a single moral dimension.
At the same time, culture should not be treated as a uniform mind. People within one country, religion or political party disagree. Gender, class, disability, generation and personal experience can change what consequences are visible. The review’s claim about cultural construction is most useful when it prompts a question about a person’s actual interpretation, rather than assigning one interpretation to everyone in a group.
Comparison also has an evidence problem. Many psychological experiments rely on educated participants from a narrow range of countries. A theory of “universal” moral cognition needs broader sampling and methods that work across languages and social settings. The review draws on anthropology as well as laboratory work to address this problem, but the underlying literature is still uneven.
Political disagreement and perceived victims
Political opponents can agree that suffering is bad yet disagree about who is endangered by a proposal. One person may foreground the people a policy intends to protect. Another may foreground those who could lose a benefit or bear an unintended cost. Their moral language may sound incompatible because each is describing a different causal story and a different set of vulnerable people.
That possibility can make a conversation more precise. Instead of asking whether the other side “cares,” ask: Who do you think is at risk? What will happen to them? How confident are you, and what evidence would change your expectation? These questions reveal empirical disagreements that a dispute over abstract values may conceal.
They also reveal where compromise may be impossible. A person’s story of harm can be sincere while their proposed policy is still unjust or ineffective. If one group is denied rights, it is not enough to say both sides fear suffering. The relevant consequences and affected people’s standing must be examined. Psychological understanding is not moral equivalence.
Some political differences may concern principles even when both parties agree on likely outcomes. A person may reject an act as a violation of a right despite believing it would produce a net benefit. A harm-based theory can offer an account of why rights feel protective, but it should not erase the person’s stated reason without evidence. This is an important competing explanation to keep in view.
What personal stories can do
Gray and Pratt suggest that stories of harm can help people recognize one another’s moral concerns. In research by Kubin, Puryear, Schein and Gray, personal experiences sometimes elicited more respect across political disagreement than factual assertions alone. The studies examined judgments and interactions in defined settings. They do not show that one story reliably changes a vote or resolves a policy dispute.
A story can communicate what an abstract claim leaves out: who lived through an event, what it cost and why the speaker regards the issue as urgent. Listening can reduce the temptation to assume an opponent has no moral concern at all. It can also help identify a practical problem that a shared policy might address.
Stories have limits and risks. A vivid example can be unrepresentative, mistaken or deliberately manipulated. Asking someone to recount trauma for the sake of persuading an audience can be exploitative. A respectful process lets people decline disclosure and does not require the most affected person to prove that their pain is real before facts are considered.
The useful sequence is to hear the experience, clarify the claimed causal chain and then examine broader evidence. Respect for a person’s account and scrutiny of a policy claim are compatible. Neither facts without context nor stories without checking are sufficient for a hard public decision.
How to test the theories fairly
A study could present a contested act and ask separately about perceived victim, severity of suffering, intention, causal responsibility, purity, loyalty and duty. It should specify before data collection which variables each theory predicts will explain the judgment. If harm ratings are collected only after a participant has declared an act wrong, the ratings may partly reflect a need to justify the earlier answer.
Stimulus selection matters. A set of bizarre but nominally harmless scenarios may not represent ordinary purity concerns; a set of obvious assaults cannot test whether harm is sufficient to explain nonharm domains. Researchers should sample across contexts and allow participants to describe the situation in their own words before imposing a category.
Measurement must work across cultures. A translated scale can preserve words while changing their social meaning. Local collaborators can help identify examples that are understandable without turning one culture into the default and another into a deviation. A large multi-country sample is useful only if the measures are comparable enough for the intended inference.
Finally, a theory about moral judgment should not silently become a theory about moral behavior. A person can condemn an act and still perform it under pressure. Institutions, incentives and power shape behavior. Measuring questionnaire responses alone cannot show how people will act when the costs are real.
What the evidence does not show
The review does not prove that all moral values are reducible to one objectively measurable harm. It argues that perceived harm provides a common psychological template. The pluralist and constructionist accounts remain contestable, and different study designs may favor different interpretations.
It also does not show that political opponents merely need to share stories to agree. Respect can improve without agreement; disagreement can remain after everyone understands the other’s fear. Dialogue may help reveal the underlying question, but policies still require evidence, rights and accountable decisions.
Nor does a psychological account authorize dismissing a minority’s concern as “only perceived.” Perception can signal a real harm overlooked by a majority. Conversely, a sincerely perceived threat can be mistaken. The task is to understand the claim and test its consequences, not to use the word “perception” as a way to avoid judgment.
Return to the two neighbors
The neighbors may disagree about facts, foreseeable consequences, rights or whom the decision will reach. Asking each to name the people they believe could be harmed may expose a shared concern and a distinct prediction. That is a better start than assuming the other person lacks morality. It is only a start: the next step is to examine evidence and the affected people’s voices.
Gray and Pratt’s review offers a way to understand how a common concern with harm can coexist with cultural and political variety. Its value lies in making disagreement more legible while preserving the difficult work of deciding which account of the world is true and which action is just.
Research references
- Gray, K., & Pratt, S. (2025). Morality in our mind and across cultures and politics. Annual Review of Psychology, 76, 663–691. View source
- Graham, J., Haidt, J., & Nosek, B. A. (2009). Liberals and conservatives rely on different sets of moral foundations. Journal of Personality and Social Psychology, 96(5), 1029–1046. View source
- Gray, K., Schein, C., & Ward, A. F. (2014). The myth of harmless wrongs in moral cognition: Automatic dyadic completion from sin to suffering. Journal of Experimental Psychology: General, 143(4), 1600–1615. View source
- Curry, O. S., Mullins, D. A., & Whitehouse, H. (2019). Is it good to cooperate? Testing the theory of morality-as-cooperation in 60 societies. Current Anthropology, 60(1), 47–69. View source
- Kubin, E., Puryear, C., Schein, C., & Gray, K. (2021). Personal experiences bridge moral and political divides better than facts. Proceedings of the National Academy of Sciences. View source